Subjects and allergen inhalation challenge
This study was approved by University of British Columbia-Providence Health Care Research Ethics Board with informed consents obtained in compliance with the local Research Ethics Boards. Seven subjects with stable, mild atopic asthma participated in the allergen inhalation challenge. Four HCs were recruited to serve as controls. All subjects were non-smokers, free of other lung diseases, and not pregnant. Diagnosis of asthma was based on the Global Initiative for Asthma criteria. Asthmatic subjects were diagnosed as mild allergic asthma, and only used intermittent short-acting bronchodilators for treatment of their asthma. Asthmatic subjects had a baseline FEV1 ≥70% of predicted, and the provocative concentration of methacholine required to produce a 20% decrease in FEV1 (PC20) was ≤16 mg/mL . The methacholine inhalation challenge was conducted as described by Cockcroft [25, 26]. Skin prick tests were used to determine allergies to cat, and the dose of cat extract for inhalation. Allergen challenges were conducted as triad visits. On the first and third days, subjects underwent methacholine challenges for assessments of airway hyperresponsiveness, and on the second day subjects underwent allergen inhalation challenges as described by O’ Byrne et al. , using extracts of cat pelt or hair. Asthmatic subjects were challenged with cat allergen to reduce confounding effects of different types of allergens. Blood was drawn immediately before (pre) and approximately 2 hours after allergen inhalation (post). Peripheral blood from HCs was collected in the morning (9 a.m.) to accommodate the time point comparing to pre-level of asthmatic subjects.
Blood collection and RNA extraction
Peripheral venous blood samples were collected into BD Vacutainer plastic EDTA tubes (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). One aliquot was processed for total and differential cell counts, and the other aliquot was frozen and stored at −80°C until RNA extraction. From thawed samples, total RNA containing miRNA was purified from 400 μL of whole blood according to manufacturer’s protocols using the RNeasy Mini Kit (Qiagen, Chatsworth, CA, USA). The yield and quality of RNA were assessed by NanoDrop 8000 Spectrophotometer (Thermo Scientific, Wilmington, DE, USA) and Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).
NanoString nCounter assay
Comprehensive assay for miRNA expression was performed using nCounter® miRNA Expression Assay Kits (NanoString Technologies, Seattle, WA, USA) at NanoString Technologies. In this method, the novel technology which enabled multiplexed direct digital counting of RNA molecules  were applied to miRNAs with some modification. Briefly, to prepare a miRNA molecule for hybridization in the nCounter assay, a proprietary DNA sequence called miRtag is ligated to the mature miRNA using a bridging oligonucleotide (bridge). The miRtags for the human miRNA are ligated and bridges are purified in one simple multiplexed reaction. After removal of the bridge, the tagged mature miRNA is then hybridized to a colour-coded reporter probe and a biotinylated capture probe. The capture probe allows the complex to be immobilized for data collection with measurement of its colour code. A total of 734 human and human-associated viral miRNAs were simultaneously assayed.
NanoString nCounter miRNA assay protocol
MiRNA assays were performed using 100 ng of total RNA following the standard nCounter miRNA Assay Protocol. Hybridizations were carried out by mixing 5 μl of each miRNA assay with 20 μl NanoString nCounter reporter probe and 5 μl capture probe (30 μl total reaction volume) and incubating the hybridizations at 65°C for 18 hours.
Preprocessing of miRNA Panel Codeset
The nCounter assay for each sample consisted of six positive controls (0.125-128 fM), eight negative controls, five control mRNAs (ACTB, B2M, GAPDH, RPL19 and RPLP0) and 734 miRNAs. Prior to normalization several probes in the codeset required background subtraction (Additional file 1: Table S3). The probes for which the background subtraction calculation produced a negative number were set to 1 for simplicity. To account for slight differences in assay efficiency (hybridization, purification, and binding) the data was normalized to the sum of 6 positive RNA spike-in controls. For each sample, the mean plus 2 times the standard deviation of the 8 negative controls was subtracted from each miRNA count in that sample. Only miRNAs with non-negative counts across all samples were retained for downstream analysis.
Technical validation using RT-qPCR
RT-qPCR was carried out using the TaqMan MiRNA Assay (Applied Biosystems, Foster City, CA) according to the manufacturer's protocol. RNA samples were measured in duplicates. The TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems) was used for the preparation of cDNA. Reverse transcription reactions were performed in a volume of 15 μl, and each reaction contained 10 ng of total RNA including miRNA. The PCR reaction mix consisted of the RT product, TaqMan 2X Universal PCR Master Mix and the appropriate 5X MicroRNA Assay Mix containing primers and probe for the miRNA of interest. All TaqMan assays were run in duplicate on an ABI Prism 7900, applying 40 PCR cycles. Ct values were calculated with the SDS software using automatic baseline settings. Ct values >35 were considered to be below the detection level of the assay. RNU44 and RNU6B were used for normalizing the expression level of selected miRNAs. The mean of Ct values was subtracted from the corresponding Ct value for the selected miRNAs resulting in the ΔCt value which was used for relative quantification of miRNA expression.
Complete blood cell count and leukocyte differentials were compared among groups using analysis of variance.Moderated robust regression in the Linear Model for Microarrays (limma) library from bioconductor was used to determine statistically significant miRNAs using a Benjamini Hochberg FDR of 1% . A p-value of 0.05 was used to determine significant changes in cell counts. Significant cell-specific miRNA expression was determined using methods as previously described . For each group, regression of miRNA expression onto the relative cell-type frequencies was used to determine the mean miRNA expression for each cell-type:
1. Multiple regression of miRNA expression onto the relative cell-type frequencies for each group g.
= 0; implies that at zero cell-type frequency there is zero miRNA expression.
; increase in miRNA expression for 1% increase in the kth cell-type frequency in group g (mean miRNA expression in the kth cell-type in group g)
g; vector of miRNA expression values for group g
kg: vector of relative cell-type frequencies for the kth cell-type for group g
2. Test statistic to determine significant changes in cell-specific miRNA expression.
Similar to the Wald test to determine significant covariates, the following test statistic was used to determine significant miRNA expression changes in the kth cell between two groups.
Test statistic comparing group 1 and group 2 for the kth
The p-value for each test-statistic was calculated by generating an empirical distribution by recalculating test-statistics after reshuffling of class labels 1000 times. A p-value of 0.05 was deemed significant.
Target prediction, gene ontology analysis and canonical pathway analysis
Target genes for miR-192 were predicted using databases, miRanda and TargetScan to list the targets identified by both. Then the target genes for miR-192 were selected out of 1595 differentially expressed genes, which were identified post-allergen inhalation challenge at an FDR of 5% by Kam et al. .
Signalling pathways and cellular processes for target genes, which were reportedly up-regulated in post-allergen challenge and predicted to be targeted by miR-192, were defined through GeneGo MetaCore databases: Functional Enrichment by Ontology and Canonical Pathway Modeling.